The Accounting Podcast

Attention: This is a machine-generated transcript. As such, there may be spelling, grammar, and accuracy errors throughout. Thank you for your understanding!

Blake Oliver: [00:00:04] If you think about it, how many people working in the government are there who probably don't want to be there and aren't productive and aren't doing anything and are just dead weight. And we'd be better off if they just took a buyout and went somewhere else where they'd be happy.

David Leary: [00:00:18] Coming to you weekly from the OnPay Recording Studio.

Blake Oliver: [00:00:24] Hey everyone, and welcome back to The Accounting Podcast, the number one podcast for accountants in the world, your weekly news roundup of everything interesting going on in the accounting profession. I'm Blake Oliver.

David Leary: [00:00:37] And I'm David Leary and Blake, I'm excited to talk to you this week. I want to talk about DeepSeek. I want to talk about Tax Dome. I want to tell you about the top ten AI accounting tools every business needs in 2025.

Blake Oliver: [00:00:48] I can't wait to hear all about that. I am eager to speak about the Accounting Today 2025 Year Ahead survey. I've got lots of thoughts about DeepSeek and tax them. That tax dome leak. The data leak. Ah, what a disaster. And my tax preparer is onboarding me on to tax Dome right now. So that was that was interesting to hear about. We also got to talk about the white House buyout offer to federal employees. That just came out a day or two ago. Um, but first let's thank our sponsors, who are our sponsors for this episode.

David Leary: [00:01:24] Today's sponsors, we have Keeper Cloud Accountant Staffing and Basil.

Blake Oliver: [00:01:29] Thank you so much for sponsoring. If you want to support us, if you want to support the podcast, please do visit our sponsors using the links that you'll hear about when we read those ads later in the episode. I also want to thank one of our listeners, Cameron Dolan, Director of accounting at the Miami Dolphins and the stadium there. Uh, I think it's the hard Rock Stadium said on LinkedIn. How do you stay updated on finance and finance and accounting news. As a busy mom of three, I know how challenging it can be to keep up with current trends. My go to strategy has been listening to podcasts during my commute. My all time favorite is the Accounting Podcast. It's an invaluable resource for staying informed and inspired. I'm always looking for new ways and helpful resources to stay on top of things. So thanks, Cameron for listening. And if you listen to the show and you want to share your love, please post it on LinkedIn. Let other accountants know about the show and tag us so that we we see your post. We really appreciate it.

David Leary: [00:02:31] And I'll refrain from making any Miami dolphin jokes. You know, since since I did not have a good week with the bills this week. So I have to.

Blake Oliver: [00:02:39] I mean, just as a bills fan in general, you're not really like in a great spot to make fun of any team, right? In general.

David Leary: [00:02:44] We made the playoffs. But the Dolphins you know, their long time arch rival from the 70s you know so it is what it is. But yeah I'm not in a great place this week. It's been a long week.

Blake Oliver: [00:02:55] I'm sorry David. Um, I guess that's. This is how I know what your mood is going to be. I should just I should really just pay attention to the games, even if I'm not, like, a fan. Just so, like I know how it's going to go with you.

David Leary: [00:03:05] Um, but my wife figured it out. This is actually good because she's like, now you're not going to waste time spending money trying to figure out how to go to New Orleans or try to figure out how to take a trip to Buffalo if they win the Super Bowl. Your life is just simpler. Now. You can go back to like, painting mailboxes and doing chores around the house.

Blake Oliver: [00:03:19] That's right. It'll be good for the business, I like that. Welcome to our, uh, live stream viewers. Don't forget you can subscribe on YouTube. Find us. We are at The Accounting Podcast on YouTube. Hit subscribe and notify and you'll get a notification when we go live, even on your phone. And you can listen in. You can let us know what you think. Tell us what stories we should be talking about. Give us your opinion. Welcome to Goat Man 1776, newly hired IRS agent here. I'm gone. Oh, no. I'm so sorry. Um, we'll talk about that. The hiring freeze, I guess. Uh, did you know, did you get your, uh, offer rescinded or did you get hired? And now you're going to take the buyout and get like, seven, eight months of salary for no work? That might be kind of nice. How you dig? Welcome to the show. How you dig or dig? You dig says. We'd love to hear you guys take on the 1099 A deduction seems crazy to me that AICPA doesn't lobby to have the Sswb limit removed. I have not read up on that pig, you dig? But I probably should, uh, maybe next episode. You know, I'm not a tax CPA, so I've tended to stay away from the technical tax discussions, but hey, I've got some good news. I can't say what the name of the show is, cause we're just doing a pilot episode, but we are working on a tax show for independent tax professionals that would dig into exactly that sort of thing. So hopefully we'll have some episodes for you on the earmark app soon.

David Leary: [00:04:51] Uh, boring accountant showed up. I was a was a little scared. I didn't see him in the comments so I'm so glad Bernie is here.

Blake Oliver: [00:04:58] Welcome. Boring accountant three coffee emojis. I've already had my three cups of coffee today and I've taken my ADHD medication, so I'm ready for the show, Ryan Taylor says. Whew! Finally made it to a live stream and light him up. Says, can we talk about Tax Dome first? I literally just bought that PM lol Fu C and another letter. Um, well, I guess we should, you know, by listener request. Okay, we'll talk about Tax Dome first. I think that's important. Um, because it's an app, it's one that if we work in tax or in outsourced accounting we have heard of, it is popular, it is growing. Um, and we need to talk about exactly what happened, uh, so that, you know, hopefully assuage some people, you know, give people an understanding like, like how this happened, that sort of thing, whether it will happen again. Um. And. Yeah. Go for it, David.

David Leary: [00:05:56] You can jump in. I'm going to add my screen share here. Hopefully everybody can see this. So this is what caught my eye, uh, earlier this week actually, Francis posted directly at Taxstone. Can you tell us more about data security incident today, where all of your users that use tags features had their data accessible to every other user on the platform via the reporting feature. And then it went on to say, can you all start taking data security seriously? This is the second data issue in a month, the first being the signature request that just what anyone with the link see the Social Security numbers on the document. And I did speak to Ashley because she, you know, she really feels like this is a disclosure issue. You know, tax professionals have to disclose where the data goes, right? And who's touching the data. If you think back like that was the major beef you had with bot keeper that time, right? They weren't disclosing it properly. That right work was going to overseas and in different hands and that type of thing. And that's Ashley's big argument. And then this exploded into a bunch of comments, and then you get other posts, for example, here, you know, people are saying urgent, there's been a data breach, protect your information. There's been a data breach at home. It's crucial to stay vigilant if your user to secure your accounts and personal information. Enable two factor authentication. These are all good things to do, but it the way based on our research and checking this out and tax domes responses as well doesn't appear to be a data breach. It just appears to be a bug. And in my.

Blake Oliver: [00:07:21] Talk.

David Leary: [00:07:21] About that.

Blake Oliver: [00:07:21] Yeah. So like what happened if I logged into Tax Dome, what did I see so that I shouldn't see?

David Leary: [00:07:28] It almost reminded me of what we discussed last week about Sage, where people would run a report and that, but they were using AI, and the AI would bring up data from other invoices that weren't yours. Same type of thing. Apparently there's a common tag many accounting firms use to bucket and tag their clients. And if you did a report on that tag you were seeing the seeing the records from other clients and other firms in yours were seeing so hourly and invoice type data.

Blake Oliver: [00:07:55] So so I go into my taxonomy client portal. I go to the reporting tool and I run a report for a particular tag. And the report pulls not just my data with that tag but also the data from other firms with that tag.

David Leary: [00:08:11] Correct.

Blake Oliver: [00:08:12] So what kind of data are we talking about?

David Leary: [00:08:14] So they're saying no individual clients were accessible. So you could view um, so no P2. So no information or documents or identifiable information PII stuff was available.

Blake Oliver: [00:08:25] Well, no it says it says let's be specific here because client details were accessible. Then, you know, there were there was information that was accessible. But they're saying it wasn't detailed information documents or personally identifiable information like social security numbers, financial accounts, client contact details or client documents. So my question is what kind of information was visible?

David Leary: [00:08:50] Yeah. And apparently it's it's like it was the invoice and the hours and like Ashley's argument and I think this is a lot of the tax professionals argument is just the fact that somebody can figure out who a client belongs to, who their accountant is, or their tax professional is sensitive information. Yes.

Blake Oliver: [00:09:10] Yeah. Okay. So I run a report and I could see invoice like here's an invoice number. Here's the hours that I invoiced that somebody else invoiced.

David Leary: [00:09:20] Yeah.

Blake Oliver: [00:09:21] It's not somebody else's screenshot.

David Leary: [00:09:23] There's not a screenshot. But so if you think about an app. Right. So the app has thousands of firms running it and they all put their tracking their own invoices that they're sending to clients and they're tagging this. And so when they run the report, instead of just seeing the data for your firm, you're seeing any other invoice that it's in the database that has that tag, right.

Blake Oliver: [00:09:44] That was it was not detailed, but they're saying like it wasn't detailed invoice information. High level.

David Leary: [00:09:50] Correct. So you couldn't like if I think going back to my QuickBooks hat, you're on a report and you can zoom in and zoom in and zoom in and eventually get to the transaction. So you couldn't. So you could see it at that summary level, but you couldn't drill down and get into it. You can't you can't actually access the record. It's basically just was just displayed on a report.

Blake Oliver: [00:10:08] But I could figure out, using that information, what the name of a client of another firm, how many hours were billed to them or something like that.

David Leary: [00:10:18] And at the end of the day, you don't want this, right? It's not you. You want your.

Blake Oliver: [00:10:23] Data. Very. It would make me very nervous if I saw that.

David Leary: [00:10:26] And so Tax Dome issued a a postmortem on this. Essentially they explained what happened. They talked about that as not there's no external access. There was not a malicious intent. Essentially it was a bug and it.

Blake Oliver: [00:10:39] Wasn't a hack.

David Leary: [00:10:40] It was not a hack. So a lot of, you know, people got very panicked about this because it's scary, right? You see, there's been a data breach and you or a data leak, and you might actually think somebody hacked in and stole data. And a lot of times it's just like, oops, we it's like very similar to the Sage thing last week. They showed data that to people that it didn't belong to.

Blake Oliver: [00:11:02] Right. And this happened. It happened over a very short amount of time. It was an hour. They're saying it was an hour on Friday morning, and 30 Taxstone users accessed the reporting function during that time, and 11 users spent more than one minute using it. So a few dozen users potentially had access to this information for an hour.

David Leary: [00:11:23] Correct. Okay.

Blake Oliver: [00:11:25] And well, let's talk about do you mind if we talk about like, how this happened?

David Leary: [00:11:29] Yeah. So how this happened is just like, oh.

Blake Oliver: [00:11:32] Before we do that, here's the here you've got on the screen right there. Exactly what fields in the time and billing reports were visible. Okay. So this was time and billing data that was tagged. So let's see if there's anything interesting in here. The billable amount, the billable hours. The amount overdue. The cost of service team members. Interest status, the team member. Revenue per hour. Duration time entry information utilization rates. So this is basically time data from other firms visible in the reporting tool.

David Leary: [00:12:07] Correct.

Blake Oliver: [00:12:08] Okay. Got it. Now this makes now this is making more sense to me. Like I don't understand why it's so complicated to explain this. Just say it was time and billing information anyway.

David Leary: [00:12:17] I mean, that's essentially what The wrap does, right? And it manages some workflows for the client and automates that. Um, so the way this happened is a lot of apps, when you build an app, even us here at earmark, right, we need to verify people's we have a text, uh, a text code we can send to people to reset their passwords. Right. So we use a third party tool to do to do that. A lot of apps, all these apps we're using are all using lots of different third party tools, because that's how you build your app faster. Like a great example. Perfect example of this is stripe. Why should every app developer figure out how to build a credit card charging thing and secure the credit card numbers and all that, when they can just get some code from stripe and drop it in? Well, take that to the next level. It's very hard to build. And I think you knew this when you were at giraffe. It's very hard to build graphs visually and bi type tools. It's very hard to build those in general. In general.

Blake Oliver: [00:13:15] Reporting is hard and it has been solved in very elegant ways. And there are third party tools that you can subscribe to that you can buy, that you can integrate into your app that just do the reporting. So you have to build it from scratch. And that is what Tax Dome was using. What was the name of it.

David Leary: [00:13:37] So they used a tool called oops I lost.

Blake Oliver: [00:13:39] But they still use it.

David Leary: [00:13:40] Yeah, they're using a tool.

Blake Oliver: [00:13:41] It doesn't matter what the name is. So it's a third party tool.

David Leary: [00:13:44] It's a third party tool.

Blake Oliver: [00:13:45] It does reporting.

David Leary: [00:13:46] It does reporting. And essentially you put this tool on and it can read. You give it access to all of your data. And there was a setting. So you think about each row. So every invoice billable these billable time invoices these.

Blake Oliver: [00:14:01] Time entries.

David Leary: [00:14:02] Are in a big table. Think about Excel right. And if somebody there's no filter on the column you could just see them all.

Blake Oliver: [00:14:09] Well and this is this is really important like because this is the big difference between SaaS products most SaaS products and like what you used to have on prem in your office. So most SaaS products are built to be multi-tenant. So everyone's data is all in shared tables. That's how taxonomy is designed. That's how a lot of apps are designed, because you don't want to have to have a table of data for every single firm. It's better just to keep it all in one massive table. And then you limit access to the data based on, like, a firm ID.

David Leary: [00:14:52] Yeah, it's filtered by some sort of firm name, firm ID, some some unique number.

Blake Oliver: [00:14:56] So, so when I, when I, when data is being displayed to me in the user interface, I only see data from that table that belongs to my firm or my firm's clients.

David Leary: [00:15:07] Correct. So this this third party reporting app has in their API security settings for row level security. And basically if this is flipped on, you only get to see the rows that are tied to your login or your you're allowed to see right that are tied to you, the end user somehow. There was a human error and this got turned off. So the row level security was turned off. And this is why all the records.

Blake Oliver: [00:15:33] And then the reporting tool would pull all the rows related to a certain tag, not just the ones filtered for my firm.

David Leary: [00:15:42] Exactly. So everybody used Excel. You have all these rows in your Excel. You turn on filters at the top, and if you're not careful to only choose the one check mark for the one piece of data. You might get way more rows on your spreadsheet than you wanted. And that's kind of the same thing that's happening here. It's definitely a human error. Um, there's some arguments going on whether or not this is regulatory reporting event. So Tax Dome is saying that it does not meet the regulatory thresholds for reporting under the following frameworks FTC safeguards rule. No personal identifiable information. Pi Phi was subject to unauthorized acquisition, so no FTC reporting is required. But I think the tax professionals and those of you that are in the live stream, please chime in on this. I think this is the big argument that's happening, as I think a lot of tax professionals feel like, yes, this was should be it does break the rule, the safeguard rule.

Blake Oliver: [00:16:34] And they and they should have to report it. Report it? Well, um, thanks for walking me through that, David. Now, this makes a lot more sense to me. And I'm trying to think, like, I mean, better QA quality assurance. Testing your code before you release it into production is probably the way to solve this in the future is right. If you if you're changing something about your reporting tool, go run reports and make sure that like this isn't happening. If somebody had just like tested it first.

David Leary: [00:17:06] They intend to do, uh, you know, steps to prevent in the future more rigorously code reviews and and testing and things like that. And you're always you're always going to say that. So the thing is like if I step back and kind of look at this on the whole. Right. Because you can see this in some of the comments. First off, let's just be very clear. This was not a hack. It appears to be a bug, but a lot of the vibe was of course this happened. It's almost like like The text. Home users were expecting it to happen.

Blake Oliver: [00:17:36] Well.

David Leary: [00:17:37] Why that I'm not fully sure on. I feel like Tax.com has a reputation problem, either good or bad, correct or incorrect. It exists, and you can tell that when people put things in the comments that says they're beyond shady. That's concerning. I've seen some shady apps in my day, you know, going through my my, I've seen apps that sent cease and desist letters to every other app on the App Store because they had like, a picker to choose a company or something off a menu on a table. Right? So I've seen some shady stuff from, from companies. But this the a lot of people are just not the fact that they were expecting this to happen or not surprised is what I find very interesting. So that's what I mean. It's a reputation problem. I poked around and I didn't see anything that was, you know, alerting to me because I've seen this in this in other places where people obfuscate what they do. And then you go into the terms of service and you're like, oh, they're using QuickBooks. That's a little tricky and shady. I don't really see that that's an issue. Um, now they do disclose like they do disclose all their subprocessors like I said, like apps use lots of third parties to build their stuff. They disclose all of them. Hold on. They got a chat.

Blake Oliver: [00:18:52] Well, the issue here wasn't the fact that they used a third party tool. It's the fact that it wasn't configured properly and that nobody caught it before they made a change.

David Leary: [00:19:00] Yeah, it flat out it was just a bug. Correct. Um, but I feel like a lot of people.

Blake Oliver: [00:19:05] That bugs like this shouldn't happen.

David Leary: [00:19:07] It's a pretty major one.

Blake Oliver: [00:19:09] Right? Right. And it's a fundamental like risk when you have everyone's data in, like, a shared table. You need to make sure that you're only showing people the data that they own.

David Leary: [00:19:19] And I.

Blake Oliver: [00:19:19] Think that's.

David Leary: [00:19:20] The vibe is people feel like Tax.com adds lots and lots of features, and they're slow to addressing bugs in general. And that's where This, this, this reputation comes from. And so people were not surprised that this happened. Um, the only thing to me that I feel is suspect at all about Tax Dome is their, uh, subscription pricing plans and their refund policy. So you cannot I always believe you should always do all SaaS on a monthly basis. Make that its software as a service. Make them earn your money every month, right. They require a one year contract, two year contract or three year subscription. That's it. And they have a zero refunds or zero credits policy, which I've seen some folks call out in reviews. So they're getting these huge commitments from firms. And then if anything goes wrong, you can't get your money back. And so that's the only thing that I think is a little suspect of like not the best practice, but everything else I've looked at like they they clearly disclose all the apps they work with, all the apps that are where they're located, where the country, where the data resides. Right. Where the. So everything seems to be disclosed. So when people say beyond shady like I question that, I'm like, I don't know what that really means. But because the only thing that I think is questionable is this locking people into these three year contracts. And then, so if anybody who got yesterday was mad or last when this breach happened, or anybody who's mad today and wants to stop using Tax Dome, you're not getting your money back. Right. And that's that that I think is not the the best policy to do that.

Blake Oliver: [00:20:59] Well I think that is a great way to end that segment. And I'm going to thank our first sponsor for this episode. Keeper by combining client communications, file review reporting and task management. Keeper has everything you need to run your bookkeeping or CAS practice. Keeper is an all in one app that allows you, your team, and your clients to easily collaborate and make your monthly close as efficient as possible. Starting with a beautiful custom branded client portal optimized for bookkeeping work. Your client can answer questions you have about uncategorized transactions, allowing you to categorize and automatically post them to QuickBooks online correctly, all without ever leaving. Keeper. Keeper is month end file review feature will surface transactions that may not be posted correctly, and with keeper's customized reports, you'll be able to increase the value that your firm provides clients by giving them reports they'll actually read. Keepers built in task management ensures nothing falls through the cracks, and it includes time tracking so you can see where your team spends their time. Keeper can also help you smooth out your 1099 season by tracking and requesting needed w-9 forms from vendors. Keeper has a very affordable and clear pricing model that starts at only $8 a month. To learn more about why thousands of bookkeepers and accountants trust keeper to manage their monthly inflows and to get 20% off your first three months, head over to The Accounting Podcast. Keeper. Okay. Let's talk about the white House buyout offer. And our listener, Goatman, 1776, said in response to my question. Got the buyout email yesterday. Nobody knows if it's legit. Fork in the road email, which is the title of this email, was the same email sent to Twitter staff and they didn't get severance. People are very concerned.

David Leary: [00:22:51] But could you clarify, was this buyout email only sent out to IRS or to lots of government agencies?

Blake Oliver: [00:22:57] Apparently this email was, uh, sent out to like every federal employee.

David Leary: [00:23:03] Every federal employee got this.

Blake Oliver: [00:23:05] Yeah, like millions of them. Over 2 million, I believe. Um, not military, not postal service. There's certain groups that are exempt from this that are not being offered.

David Leary: [00:23:15] Might have guns. They don't want to get.

Blake Oliver: [00:23:17] Riled up.

David Leary: [00:23:18] I get it.

Blake Oliver: [00:23:19] Um, so you can you can read the email, uh, on pm.gov/fork and it's pretty simple. Basically employees, all they have to do is send an email to HR at opengov from their government account and type the word, resign into the subject line of the email and hit send. And that is your acceptance that indicates your acceptance of this offer. Uh, and so, uh, you know, what is what is the offer basically. Uh, if you do that by February 6th and this includes, I want to say, IRS employees, right. That's the tie in here to tax season. And accounting is if you work at the IRS and you want to take this buyout offer, you just got to do that. By February 6th, you will get 7 to 8 months full pay and benefits. It's through September 30th. Um, there's limits on how much the federal government can pay employees for buyouts, but the administration has figured out a workaround. They're just going to put everybody who takes this offer on paid administrative leave for 7 to 8 months. So it's effectively the same thing. So you don't have to come to the office. You don't have to work. You're on paid administrative leave until September 30th. And then after that you are unemployed. Now, if you don't accept the offer, you've got to expect to return to the office five days a week. Because that is is a something the administration has said they're going to do for all federal employees. And the white House expects 5 to 10% of federal workers to accept this buyout offer. So the question is, what does this mean for the IRS? Well, the IRS has had its hiring frozen since January when the administration froze all federal hiring.

Blake Oliver: [00:25:19] And they have a natural 8.5% attrition rate. That's an average over the last three years. So take 8.5%. And then, you know, maybe all of those people are going to just take the the buyout anyway. But I would expect that some people who weren't going to leave will now leave. So how much could it be? Well, let's say it's another 10% at the IRS, on top of the 8.5% that we're already going to leave. Now you're talking like when you're edging towards like 20% of the IRS workforce. And I mean, it could be more than 10%. Somebody on Twitter was saying in response to this, hey, you know, like Iris isn't the best place to work at right now. Maybe more will take it. And if you know what you're doing in tax, you can get a job. So I mean, so what does this mean for the IRS? It means we're going to have a lot fewer people probably. I mean, my guess, my gut saying 10% or more of the IRS workforce will resign, take this buyout. So what is it going to happen? Are you going to have longer wait times, slower processing, just like in Covid. Remember those good times. And then depending on, you know, what side of the fence you're on, reduced audits which, you know, that might be great, good in your opinion, or it might not be good. Um, but that's that's the buyout. That's what's going on.

David Leary: [00:26:42] See something from Trump in the last week and a half here or so, where he suggested that the 80,000 new IRS agents, they can get jobs, but they're going to have to work the border instead.

Blake Oliver: [00:26:53] Yeah. Is that you know, um, I'm going to take that as one of those statements. That is a joke and not going to become policy. But yeah, it would be pretty funny if they, uh, took the armed IRS agents and sent them to the border and made them do border enforcement, wouldn't it? Oh, man. Yeah. Juventino says that everyone got the email. Boring. Accountant says nice. After you resign, reapply with a hiring bonus. Tino says, I've been fielding questions all morning long with our management team from our staff. That's not necessarily true. We need to be careful with misinformation. Still trying to digest everything. Well, yeah, I guess. I mean, the email went out, but as we saw with Twitter, not every employee did get a buyout. So if you reply to that email and you say resign, you know, maybe you'll just get fired. Like, who knows what will really happen, right? So it's a good point, Tino.

David Leary: [00:27:52] And then just to make sure I read the comments, this is the subject line was exactly what they sent out for Twitter.

Blake Oliver: [00:27:59] Same. Yeah. Same email subject line that the Twitter employees got when Elon Musk did the same thing. So this is clearly a doge.

David Leary: [00:28:05] A Doge.

Blake Oliver: [00:28:06] Yeah, a Doge move. Doge action. Um, and, you know, it'll probably be very effective. Like, uh, it's a great way, actually, if you want to reduce the size of the federal government is a is a great way to do it. And if people take it voluntarily, it's kind of hard for anyone else to, like, complain, right? Um, and it will save us a lot of money. And if you think about it, how many people working in the government are there who probably don't want to be there and aren't productive and aren't doing anything and are just dead weight. And we'd be better off if they just took a buyout and went somewhere else where they'd be happy.

David Leary: [00:28:39] Yeah, if you're burnt out and you're just just mailing it in anyways, you're not really working hard. I could see, yeah, this is an easy way to start to trim. I could see it.

Blake Oliver: [00:28:50] And it's another example of how return to office mandates are actually really just a way to get people to quit. Quit so that you don't have to fire them and pay them, uh, unemployment or whatever. Um, POTUS Goatman says POTUS also said at a Nevada rally that they are working on a plan to terminate newly hired agents. Um, yeah, I, I guess if you already got your job at the IRS before the hiring freeze, you got to keep it. But anyone who had an offer that was still out there that hadn't gone through like that all got rescinded.

David Leary: [00:29:23] They're rescinding.

Blake Oliver: [00:29:23] This. We'll see if the newly hired people get fired, too. The problem with all of this is, as we know, for every dollar you spend on the IRS, the IRS gives back like 5 or $6. So, you know, you put $1 million into the IRS in funding. You get 5 to $6 million back in taxes that otherwise wouldn't have been collected because enforcement is so low right now. So, you know, if you want to solve the budget crisis, if you want to reduce the deficit, then cutting IRS funding is not necessarily the way to do it. And, you know, the IRS may not be one of those places where you want to trim the workforce so much. I don't know. But again, there's part a part of me that says, well, there's always that like 10% that is just dead weight in an organization.

David Leary: [00:30:18] Corporations and and I think that's unfortunately or fortunately, whatever. Elon and Doge has proof they can point at Twitter. Look, we got rid of all this fat and Twitter still runs questionably, but it runs. It didn't crash. You're not getting the fail whale from back in the day. So they have data to point at this. And and I think I heard an argument I made on the All In podcast about an extinction event. You you basically are going to cut so much and then whatever these government services survive. Yeah, those must be the ones that are valuable and keep those ones only.

Blake Oliver: [00:30:53] Well, and if the people who walk out the door are the people with all that institutional knowledge, that's the big concern is, oh, all these really experienced IRS people are going to leave and they have all this knowledge, and now you'll be left with a bunch of inexperienced people. Well, that could be a bad thing, but that could also be a good thing, because it's the people who have been there the longest who are the most resistant to change. And if you want the IRS to actually change, people who have been there for like 20, 30, 40 years are not the ones to do it. It's the new people. So, you know, we actually saw this with DeepSeek. I think we should pivot to that. David, the new AI model out of China, because one of the fascinating things about Deep Seek is that apparently the guy who did this, who created Deep Seek, the ChatGPT killer, uh, did it hiring people, like, fresh out of college with, like, 1 or 2 years of experience.

David Leary: [00:31:49] Sometimes rewind a little bit on this. Just explain for those who have missed the news. So on Monday, Deep Seek took over over rival OpenAI's ChatGPT as the most downloaded free app on the US Apple Apps App Store for the Apple App Store. So out of nowhere, this app just got momentum and it took over the over the number one spot. And what happened? It actually wiped out billions in market cap. Shares of Nvidia fell nearly 17% on Monday. The Nasdaq dropped more than 3%. Some people are calling this our Sputnik moment, where the US is in this technology and space race with the Russians. And then one day, all of America had to watch Sputnik fly over. This is kind of the same thing. Everybody in the US, they just what's the big announcement they had last week? The half $1 trillion with governments going to pitch in with open AI and all that.

Blake Oliver: [00:32:39] And the reason deep tech was such a big deal is because the company is saying that they only spent only $5.6 million to train this model, and OpenAI spent something like $100 million to train their latest, like four. I think it was ChatGPT four model. So like that is way cheaper. And they did it using really innovative Techniques that saved a lot of processing power, so they didn't have to use so many compute cycles. They didn't even need the most advanced chips they had, like Nvidia chips that were not as good, not as many. And they basically for a fraction of the cost, built the same thing as open AI. And so the markets freaked out because they're saying, well, you know, how valuable is Nvidia really, if you don't actually need these super advanced chips to build really sophisticated AI and the companies down like 15%, it's the largest wipeout of market cap in the history of the stock market, like $600 billion. Just like evaporates in market cap.

David Leary: [00:33:46] In an afternoon, half an afternoon. Yeah. And what's interesting is like, we caused this, right? We had these embargoes and we stopped letting them get access to the fastest chips, thinking like, hey, if they don't have access to the chips, China can't compete with us on AI. So we tried to rig the game and all it did was create one of those necessities the mother of invention, invention moments where because they were so starved for the computer power, they had to figure out how to write the code more efficient. So the engineers had to be better than us. We basically created a monster over here. We're worried about. We're like, literally, we're worried about whether or not we have TikTok in this country. In the meantime, the real threat from China just happened. And when I say threat, it wiped out 3% of the market value of the Nasdaq. That's 3% from all of Americans. 401 (K). That's the threat.

Blake Oliver: [00:34:35] It's, uh. It's funny how it backfired, right? Yes. We restricted the export of the most advanced Nvidia chips to China. Now, Nvidia apparently did something that sort of worked around that and gave them chips that were like almost as good, but they couldn't get that many. They were definitely at a huge disadvantage. And so that constraint, not having access to lots of chips and not having access to the best ones, forced the Chinese researchers to figure out how to do more with less. So, like you said, David, we created this situation. Um, it's it's funny. It's like I remember as a little kid right at the end of the Soviet Union, right? Like, I remember how, like, like America it was. It was amazing how the Soviets always managed to do so good despite their limited resources compared to us. And it's that, like, competitive spirit and limited resources that make people excel. Um, and sometimes when you have too many resources, you get lazy, you get fat, you get slow. And I think that's maybe what happened with us. Right. We just said, oh, we can we can keep improving models just by spending money. And OpenAI has access to basically unlimited money because money was cheap. They're getting massive investment. They can just blow money on on building these models. But you don't actually need to do that.

David Leary: [00:36:00] And this might change the game on everything. The electricity that's needed to run I basically Walmart at I, which could be good for all of us in the country. We might be able to get I for a lot cheaper now possibly.

Blake Oliver: [00:36:12] I want to say to side note, um, if you want to try the deep seq model, it's an open source model and you can actually try it via perplexity. I so I went into my perplexity, I count I don't know if this is available to all subscribers, but I'm on the pro account like I pay 20 bucks a month. And when I went in the other day yesterday, I think it was, I saw that I could toggle in a dropdown, I could toggle to use the new deep seq R1 open source model. Now it's not in China, it's a version. It's a it's a clone of that model hosted in the in the US. And its reasoning is incredible because perplexity actually shows you the reasoning steps that it goes through. And just looking at, oh, one's reasoning In ChatGPT the reasoning flow versus R1. R1 seems to be doing a lot more reasoning, and that's how it is so good, even though the the model was is smaller. I don't know if it's actually smaller, but there's something about like the way that they train these models is they didn't have to train it as long because of the way that it reasons. So it like divides up tasks. Oh, here's the here's the analogy I heard. That's good. So OpenAI one, it's like having access to a librarian who's read every book. So you go to that librarian and you ask a question, and the librarian goes into, you know, her knowledge and and gets you the answer. Well, the R1 model was not trained that way. You go and you ask the librarian and then she goes and like looks up the answer using reasoning. And then they've developed specialists specialist Uh, agents that know certain topics, so it farms out the work to smaller models, if that makes sense. Rather than having one big model.

David Leary: [00:38:10] I think. And I could be completely wrong, and I might this might be engineering folklore. But I think in a weird way, that's kind of how Google attacks search, because to some extent the Google search box is almost instantaneous. And so instead of having Google search search for everything based on the words you start typing in, but they would set up servers for just R and server for S, and so it would route that traffic and it would only have to search within its constraint. And that made everything faster right. Like and this could be engineering folklore. I might not know what I'm talking about at all. And people can cancel me for this. But back to my understanding. This you, you divide up the load so you don't have to have one thing. Try to search everything and just be really smart.

Blake Oliver: [00:38:50] That is. Yeah, we may not be using the correct technical terminology, but essentially that's what seems to have happened here, is that DeepSeek figured out how to divide up the job into smaller experts.

David Leary: [00:39:03] And if it's smaller, you could have weaker processes do the work.

Blake Oliver: [00:39:05] Exactly. Um, Martin brings up a good point in the live stream. Good to know that development costs figure is coming from a Chinese company. Trusting them has always worked out so well. So, Martin, I'm really glad you brought this up because I had the same thought I when I saw the market crash. Well, it didn't crash. Dip. Uh, when I saw the market dip, I thought, where is this information coming from? And there is no independent third party verification of any of this financial information. It's in the research report that came out from this company, but nobody has verified it. So think about that. A figure that is not independently verified, not audited, caused the caused the market to dip. How much? David. Like the 3%.

David Leary: [00:39:50] 3%.

Blake Oliver: [00:39:51] Yeah I mean and so if you were like a short seller and you wanted to profit off of this stuff, what would you do? You know, you'd take a bunch of short positions, issue the report and then close them out. Make a lot of money really fast. Like there's something that's a good problem there.

David Leary: [00:40:08] And I think I saw a comment that regardless of the specific number. So let's say the Chinese government subsidized some of this and put some money in. And they're not reporting that just the way it's done. It still was probably cheaper than any of the AI models we built. And not that the bigger thing for me is it's not so much the building of it. Like, yes, things cost money to build. It's the running. So if you can and I'll post a link in our show notes. I don't want to put it in the stream because we'll lose everybody that's watching us to go watch this YouTube video. But there's a guy in his garage just running on his PC. He has a PC, he's just running it. He downloaded the open source and he's running it on a PC in his garage. Yes, he has an Nvidia graphics card. But essentially, you know, you can run this on your own machine in your garage. That's where it's at. And you couldn't do that with some of these other models. Not at the same level.

Blake Oliver: [00:40:55] I'm curious exactly what the cost difference is on the ongoing use, because that could have significant implications for businesses that are relying on AI like ours now. Um, so I posted on Twitter about how we use AI now to write the first draft of every CP course that goes on to earmark. That used to be a human process that took 3 to 4 hours, and now we do it for 2 to $3 of compute on open AI using their, uh, Foro model.

David Leary: [00:41:24] So what if that could be now $0.25 or $0.15 for us, right. As a business owner, we're going to switch to that. That's a threat to all the US companies.

Blake Oliver: [00:41:33] It is.

David Leary: [00:41:34] It's like shopping at Walmart.

Blake Oliver: [00:41:35] And the models, the models open source. So I could just switch all of our processes to some company that's hosting the open source model. Or we could even buy hardware to run it ourselves and just pay for electricity. Right? So, um, it's looking like I mean, I'm doing a perplexity search right now, so I haven't been able to verify that this is correct yet. But perplexity is telling me that the API cost per million tokens for deep sector one is like inputs are $0.14 to $0.55, and outputs are like a little over $2 $2.19. That's per million tokens. And ChatGPT one Pro is inputs are $15 per million and outputs are $60 per million. So we're talking about like on the input side I mean that's a that's you know, could be a difference of like 100 times cheaper.

David Leary: [00:42:36] And that's and in a way it makes it look how ridiculous it is. The earlier announcement this week, the half $1 trillion the government's going to support the AI czar or whatever. I don't even know what the announcement was, but it's a lot.

Blake Oliver: [00:42:49] It's right.

David Leary: [00:42:50] Half $1 trillion to be spent to protect are our lead in AI. It's obviously not how much you spend. It's about how efficient you get at it.

Blake Oliver: [00:43:00] So here's my tie in to accounting. David. Um, so when you have a lot of money, it can often be hard to innovate. If you have too many resources, if you have too much money, you waste it. But if you are constrained, you have to be smart. It causes you to be more innovative. And I think this is why we never see large firms innovate. Large accounting firms cannot innovate because they have just too much money. So same thing with large associations. Um, you know, a great example is like, uh, you know, large associations that try to build their own software, like the AICPA is building like audit software. For years and years, we've heard about this amazing dynamic audit solution. Like, I never hear anyone talk about this thing, right? They just pour money into it. But small technology firms can go and build like audit solutions that are like amazing. They get adopted. Same thing with small firms. Small accounting firms can change processes, can try new things, can do innovative billing models. And they have to because they don't have money to waste. So I think it's the big firms are just too profitable. That's why they can't innovate. And and so I don't know if there's a lesson here for like technology companies or firms. But like if you really want to innovate you've got to create a you've got to create some sort of restriction. You know, if you want to be creative, you can't have like everything possible available to you. You have to limit your possibilities.

David Leary: [00:44:36] I went through this at Intuit when I built View My Paycheck because it was really grassroots, and we went up without budget approval, without any of that. And that's why eventually got built and got out the door versus the other way, which would have been create a PowerPoint slide. Go ask for ask for budget money. Build the team. Go build the thing. Because we're a resource constraint, we're actually able to get it out the door because we just had to do it. What's the cheapest minimal way to do this to get it done? And it really a lot of it's just management. I think when you have more money, you have if you're going to have a half $1 trillion, you're going to have a lot of managers managing that money. And if you think about all these government, you look at the the Dodge layoff, the fork in the road, right. All these government agencies have budgets of billions, if not trillions. Some of them I think I heard FDA is like $1.5 trillion budget. Yeah, right. And they have these huge budgets. When you have that much budget, you have to have lots of management. And management really gets in the way of actual innovation at the end of the day.

Blake Oliver: [00:45:34] So here's my idea for how we can fix government, specifically the IRS, using what we just talked about. And that is instead of trying to change the existing IRS, we actually do something similar to what Donald Trump has proposed. President Trump has proposed creating a new revenue service. He calls it the external revenue service that would collect tariffs and all that sort of thing. But I think we should think bigger than that. We should create a new revenue service that is completely separate from the existing IRS, and that is where you put the innovative people who want to actually build the systems and the processes to handle revenue for this country in a way that is fair, efficient, effective, not painful, easy for tax pros to work with. You could do this with a few thousand people, maybe even a few hundred people to start, if you have the right tech and you can start again, like and and there's ways to transition people over. You could take the easiest tax returns and have the, you know, new revenue service, start doing those, or actually maybe even take some of the most complicated, least frequent returns, the rarest kind, and just shuffle them all over to the new revenue service to deal with, and they can figure out how to optimize that.

David Leary: [00:47:02] And this could have been delivered in the letter. You can take this buyout or if you'd like, you could go to the new startup.

Blake Oliver: [00:47:10] New revenue.

David Leary: [00:47:10] Service. Right. The new the IRS 2.0 that we're going to build from scratch. You can join that team and you're actually going to get a personality that is not that maybe likes the risk of that, or it sounds good and not take the buyout and.

Blake Oliver: [00:47:24] Bring in people from tech to work in the new revenue service to, to build the systems, you know, bring in people from into it, bring in people from, uh, from Oracle NetSuite. Right. Bring in people from all these different apps in this ecosystem that know how money moves. Bring in fintech people and and, you know, bring in, of course, tax experts. And existing IRS people and combine them all together. The best of the best. And you could build a new IRS that needs 20%, 10% of the people that an IRS needs now because you're using technology to process all this information. I mean, people are literally still sitting in the IRS keying in information from paper tax returns. This does not need to happen. And it's insane that it's still happening. And if you start from scratch, you could just eliminate that from the get go. Uh, so, David, I think we should thank our next sponsor.

David Leary: [00:48:23] Yeah, I'll jump in on that here.

Blake Oliver: [00:48:25] It is going to be cloud accountant Staffing.

David Leary: [00:48:28] And I'll read it if you want to add the banner here.

Blake Oliver: [00:48:30] If you need if you need help, check out cloud Account and Staffing.

David Leary: [00:48:36] I need that. So in case you've missed the last hundred or so episodes, Blake and I have been discussing almost weekly that there's an accountants labor shortage. Regardless of the root cause, the problem is real. My social media feeds are full of firms attempting to fill open positions on their teams, but how can anybody increase their staff size if there's a labor shortage? How do you actually hire people? That's where cloud accounting staffing comes in. They'll help you hire a full time team members for your firm that reside in the Philippines. How much should your firm change? Or for that matter, your life? If you could add 40, 80, or 120 hours of capacity to your firm in 2025? Cloud Accountant Staffing was founded by a firm owner who grew his firm using offshore talent, and now he's applying everything he learned to help you grow your firm. If your firm is in need of expert bookkeepers, accountants, CPAs or virtual assistants, head over to The Accounting Podcast. That is The Accounting Podcast Karabaic. Thank you. Cloud accounting. Staffing.

Blake Oliver: [00:49:30] Let's talk about accounting. Today's 2025 Year Ahead survey. I've got a list of seven stats from this survey that I really enjoyed.

David Leary: [00:49:41] Now who do they do they survey for this?

Blake Oliver: [00:49:44] Oh that's great, I wish I had that in front of me. Uh, let's see, they surveyed. I have to I have to open the full report. That means I have to download the report again.

David Leary: [00:49:55] Oh, you got to put it in email. That's okay.

Blake Oliver: [00:49:57] We'll see. Oh, wait, I got it, I got it. Uh, where's the key? Where's the methodology? Okay, they surveyed 291 accounting professionals and 65% were CPA firms, 19% non CPA accounting firms, and then 16% tax only firms. So it skews, uh, towards CPA firms. And there's a broad range of sizes skewing more towards the under 20, which is you know most of the profession is uh firms under that number. Okay. So what are the key takeaways. So you don't have to read it yourself. Let's talk about revenue growth. 69% of firms expect revenue growth of at least 4%. I would hope they would have revenue growth of 4%, because that's only a little bit more than the inflation rate. But 24%. So almost a quarter anticipate double digit growth 10% or more. What about advisory work? We keep hearing about how important advisory work is. Um, we're actually spending less time on advisory work than last year. So only 31% of firms, um, this year, 31% of firms spent less time on it this year.

David Leary: [00:51:22] Are they getting more?

Blake Oliver: [00:51:23] They spent more on compliance work.

David Leary: [00:51:26] Are they more efficient at the advisory or is it just there's so much compliance work.

Blake Oliver: [00:51:29] I think there's just so much compliance work. And also, you know, my feeling is that if you do compliance work well, it's very profitable and it's very scalable. And so, um, while there's nothing wrong, in my opinion, with doing an advisory, it's a great business. Um, like, if I were building a firm again, I would be focusing on compliance services. Uh. All right. So how about hiring? Half of firms are going to hire in 2025, but they're only going to hire half as many people compared to 2024. And when it comes to remote employees, 62% of firms refuse to hire any remote employees like fully remote employees.

David Leary: [00:52:15] So more than half of all for all people surveyed says their firm will not hire remote employees.

Blake Oliver: [00:52:20] Yeah, and that's a 10% increase over last year. So we are moving away from remote work. Now, that still means that 38% of firms will consider hiring a remote, a fully remote employee. Um, and actually, while we're on that, What is the breakdown of remote versus hybrid versus in-office in public accounting? So it's 56% that work full time in the office. So a little over half work full time in the office all the time. All the employees. And that's an increase of 4% from last year. About a quarter of firms 24% have hybrid arrangements. And then another 20% have fully remote. But that's a decrease. It used to be 25%. So we're down to 20% of accounting firms in this country or surveyed by accounting today are fully remote.

David Leary: [00:53:20] And you see that's down 5% from the year before. And I imagine that was down 10% from the year before and 20% for the year before because we were in Covid times. Right. Like is this trending to zero?

Blake Oliver: [00:53:31] I don't think so. I think we're just normalizing. That's my feel because before Covid it was like less than 10% of firms were fully remote. And we doubled that number essentially, which is a lot. So I would just say this is like going, you know, normalizing back to the trend. Um, but, you know, if you're a firm that hires fully remote, like there's a lot of people who want those jobs, so you actually have an advantage compared to firms that require in office. But we also saw from the earlier data point that firms are hiring fewer people compared to last year. So if you need fewer people, you can be pickier about where they work. Basically supply and demand kind of situation, right?

David Leary: [00:54:15] Yeah.

Blake Oliver: [00:54:17] Let's talk about it. Spending. Tech spending. Over half of firms are going to increase their IT spending in 2025. It's actually 60% and that's up 10% in previous years. And get this technology as a percentage of the firm's budget. It. So percentage of expenses is what I'm taking this to mean. It's 21% now. 21% of your firm's budget goes to tech. That's way higher than when I had my firm. Even just well, that was in like 2014, 2015 when I sold that. I mean that's amazing. And larger firms spend even more 30% or more of their budget is is it and technology. Software is eating the world, as Marc Andreessen said. Yeah. It sounds like you have some thoughts on that.

David Leary: [00:55:10] I'm wondering how much of that is because a we're adopting new technology, or is it just a lot of the apps, once they know they're established in a firm, they start raising the prices. Maybe it's a mix of both.

Blake Oliver: [00:55:21] Price inflation. Well, you would think though that like if if like you've always said, David, whenever QuickBooks raises its prices, that should be your trigger to raise your prices. Prices in your firm. So then if if your prices are going up along with the tech spend, you would expect it to stay a similar ratio, a similar proportion, but it's actually growing. And that to me indicates that firms are figuring out how to do more work with tech and less with people. And, um, yeah. So like, it's just it's flipping from cost of labor to software subscriptions and that's great for the accounting technology industry. Um, and it helps to solve the talent crisis. How else are firms going to keep up if they can't hire people, they're going to buy software to do the work faster.

David Leary: [00:56:11] And finally figure out how to do it with tech. That promise of replacing bodies.

Blake Oliver: [00:56:16] Allison says. Why fewer people? Is that because firms are taking on less work, or just that they have been trying to find people in 2024 and had a tough time? I don't know, maybe it could be a lot of things. Uh, could be that like we just talked about technology means they don't have to hire as many people. Maybe AI is starting to have an impact on firms. Um, maybe everybody just hired a lot of people last year for whatever reason, and now we're just naturally pulling back. I'd be curious to know the answers. Um, and if you're listening, if you're on YouTube or on LinkedIn with us as we stream here, let us know. Like what? What is your firm hiring fewer people? Are you hiring more and if so, why? Okay, last stat still on the technology theme. How many firms are using AI tools? It's 23% of firms right now that responded to the survey. 23% of firms are using AI tools. Um, which is kind of like what I would expect, actually, although it's pretty much the same as when I surveyed my, uh, speaking conference attendees, like when I went and did speaking sessions and I surveyed the live audience and a lot of these conferences with a lot of CPA firm attendees. I would only get maybe 20% of the audience saying that they'd actually used AI. Now it's only 23% using AI tools, but a third are exploring the technologies. So add those two together and you get half of firms are either using them now or exploring using them. So we should see rapid growth in AI adoption in firms over the next few years. All right. That is all of my stats from the Accounting Today Year Ahead survey.

David Leary: [00:58:06] I think this will go off of this a little bit. So I saw an article in The Accountancy Age and it said the top ten AI accounting Tools Every Business Needs in 2025. So before I display this table, Blake, what kind of things would you expect to see on this table if it's the top ten AI accounting tools?

Blake Oliver: [00:58:25] Uh, top ten AI accounting, tax research, maybe would be would be one, like a tax research tool. Uh, I would I mean, I expect to see like the chatbots like ChatGPT and Claude and Perplexity.

David Leary: [00:58:38] Or maybe like all the hot new accounting AI agent startups that are happening. Uh, the basis and, you know, even all these ones that are out there, maybe you'd.

Blake Oliver: [00:58:46] Want to see those. I mean, if this is for like a general business audience, then yeah.

David Leary: [00:58:50] So let me open this up for you. I'll add this to the stage and here's the here's the table. The ten best AI Tools to Know in 2025 for businesses to know. It's just a list of goals. Basically, it's QuickBooks, Zoho, Xero this.

Blake Oliver: [00:59:06] Content.

David Leary: [00:59:07] I this I couldn't figure out. It might be.

Blake Oliver: [00:59:09] It's an accountancy age. They don't tend to do that much of that. I mean they.

David Leary: [00:59:13] Usually positive it's sponsored content or not, but it's it's a lot of the AI that's already in the GLS they're talking about. Um, even FreshBooks is on their NetSuite. So there are a couple apps like Vicci is on here, which uses AI to scan bills and those types of things, and then trillions on here, which will scan your lease contracts and all of that. But in general, I was shocked to see an article with that title and just see a bunch of girls like, where are, like you said, where are the where's deep sea? Where's ChatGPT? Like, how are they? How are the basic even just clawed to do your emails?

Blake Oliver: [00:59:51] Yeah.

David Leary: [00:59:51] They just don't exist. It's it's. You're right. Maybe this is a paid article.

Blake Oliver: [00:59:56] Um, it kind of annoys me.

David Leary: [00:59:57] That.

Blake Oliver: [00:59:58] I don't know, whoever did this does not know what they're talking about. I'm sorry, but because the AI tools that are built into these GLS are really mediocre right now. I don't know, maybe our listeners have had a different experience, but like I am not impressed. And zero is on this list. So I'm going to go ahead and rant a little bit about zero. I'm going to complain as a zero user because zero just unveiled their new tagline. They've changed their tagline from Beautiful Accounting software to your business supercharged, charged, which I guess means that they're going to use AI to like, supercharge businesses. But when I go into Xero and I reconcile, it's still the same flow that I have experienced for like ten years now, more than ten years, it has not really meaningfully changed. And they could be using AI to automate the categorization of all the transactions. And it sucks. It's terrible. Whatever they did was with machine learning a while ago and it does not get it right. Like bank statement line comes in. This is something that if I gave it to ChatGPT, even like the mini model, like the cheapest model with like a GL sample and I just gave it bank statement lines, it could, it could pick the vendor and it could pick the account code like 99% of the time. And Xero doesn't I don't get it. Like why are they not investing in this? And why is I don't know if this does this work well in QuickBooks either. Like does the automatic suggested transaction categorization work that well?

David Leary: [01:01:28] I mean, they made some changes. It's getting better. But I also like went to the high level business feed the other day, and I try to upload a bill to the high level AI part that's inside of QuickBooks. And then the bill just never showed up. So then today I had to go to the enter bill's screen, upload a bill so it would upload it, and then do the AI reading on it. So there's a disconnect on the global AI in QuickBooks and the specific one. Um, but this ties back to deep tech, right? It's probably very expensive for Intuit and Xero etc. to lease this compute power. I mean, add this stuff to their thing.

Blake Oliver: [01:02:04] I don't buy that though, David, because like ChatGPT mini four Pro Mini is insanely cheap. We use it for like pennies per week.

David Leary: [01:02:15] We can afford it.

Blake Oliver: [01:02:16] It's true. If we can afford it, then like Xero and Intuit can afford this stuff. And I mean, it's really as simple as just like you run a prompt and you can if any product people are listening, I'm happy to explain how to do this to you. For you is is you. Periodically you save like a copy of the GL. You send that over to like an AI agent that's dedicated to that company file. Okay. And so it has the it has like the last I don't know, x months, years, maybe the whole thing of GL data for that bank account or maybe even the whole company. Just give it the whole GL. Right. Maybe you can do that. And it has that in its knowledge. And then you just run a prompt for every single transaction that comes through on the bank feed as to like, okay, based on this data that we got through from the bank feed, what should the vendor be and what should the chart of account be? Look at the GL to figure it out. These mini models can do this with insane accuracy. Like this is solvable. I could go to Intuit or zero and do this like I don't understand why they haven't figured out how to do this.

David Leary: [01:03:25] It feels like the models right now for bank feed matching, just look at amounts first, I think, and then that's it because it's easy to match the amounts. And then that's why I always suggest the wrong ones. So like it'll suggest one that you haven't matched yet. That's four weeks away. And like well obviously that's not it. It's this one. It actually it says Starbucks like like this is the this is the one.

Blake Oliver: [01:03:44] I, you know, I think it is I think it's that all the product people like boring accountant here says in the chat that they're not focused on solving these user experience issues, because this is not the sexy new feature that brings in a bunch of users that marketing and sales can use to sell the software, so they don't focus on it because they're not actually focused on like, what are people doing every day that we can make easier in these products? And like, that's the big problem when a company gets really big is they stop paying attention to the day to day users. So yeah.

David Leary: [01:04:19] Ultimately, where I saw a tweet about this, about product managers that don't want to talk to customers ultimately or get in the weeds and test stuff out. But this ties back to this, I think, people's frustration with Tax Dome. They feel like Tax Dome is adding all these new things and not fixing the unsexy, boring stuff, right? It's always about the new feature. The latest thing I can sell your firm the hottest new thing. And yeah, people don't ever want to do the boring, the boring work that adds up over time to be a super valuable product.

Blake Oliver: [01:04:52] Well, and the attitude of move fast and break things does not work in the accounting profession. Yes, we need to know that our client data is secure, that you are making that a priority. And so issues like this, they seem small, but I guarantee you that Tax Dome is going to take a huge hit on their on their sales because of it, because they will have lost the confidence of a lot of lot of people as a result, whether that's. Warranted or not, it's the appearance of not having the security that is just as much a problem for these tech companies.

David Leary: [01:05:32] They're not having it.

Blake Oliver: [01:05:33] Yep. All right, David, we should thank our final sponsor. I'll go ahead and read this one if you want to put up the banner. Thank you to Bazel for sponsoring this episode. Is your accounting firm looking for a better practice management solution? Are you tired of juggling multiple apps just to get work done? Well, you need to hear about Bazel. It's an all in one practice management platform that finally brings together everything under one roof. Client portals, e-signatures, tasks, workflows, calendars, invoicing, you name it. It's everything you need to run your firm smoothly. You can deal with just one product, one user interface, and one subscription. And plus, it's powered by Amazon Web Services, so you know your data is secure. Occur. What makes Bazel unique is its beautiful, clean design. You'll feel right at home because it works like your favorite apps. There's no complicated training or lengthy setup. You can be up and running in minutes and get this. You'll receive complimentary onboarding services, and if you ever need a hand, there are world class support team. Is there 24 over seven with Bazel? There are no hidden fees, no surprises, just straightforward pricing. That makes sense. The monthly cost is a low $30 per team member, and that includes unlimited clients, allowing you to scale your firm without huge software bills piling up. If you're ready to transform your practice and want to give Bazel a try for free, head over to The Accounting Podcast Bazel. That's The Accounting Podcast dot promo. Bazil. Tyler says in the chat on LinkedIn. Amen, Blake move fast and break things is how you fail. Thank you. Yeah, it it works in certain areas of tech, but not in areas where you are partnering with professionals for whom move fast and break things does not work.

David Leary: [01:07:21] And I think this applies to firms, right? I think firms like to put in reinvent, to tear down the walls, rebuild type huge initiatives. And a lot of times it's just like polish the stone, just fix these little things along the way and they'll all add up to a big change eventually. But they're just you're right, it's not sexy, it's not exciting. And so these small things just start to keep getting worse. And the next thing you know, you have tons of pebbles that are just hitting you like a big huge stone. Now firms. Yeah.

Blake Oliver: [01:07:50] Nobody wants.

David Leary: [01:07:51] To fix the broken stuff.

Blake Oliver: [01:07:53] Right? It takes me one second or 2 seconds or 3 seconds to code that transaction, but that adds up day after day. And it just feels pointless because I know that it could be automated. So it feels worse now. Um, you know, there's like, there's a part of me that just wants to, like, build a browser plugin that does this right. That just looks at zero. I mean, maybe we'll get that with ChatGPT tasks featured or the cloud computer use is, you know, I'll be able to say, just go into my GL, go into my accounting software every day and categorize all the transactions. You know, how to categorize and leave notes on stuff that you're not sure about. I mean.

David Leary: [01:08:34] And look what I did yesterday and the day before. Yeah. So if these guys.

Blake Oliver: [01:08:39] If these software companies don't move fast enough, I guess eventually we will just get these like agents that can access a web browser and that can do this stuff for us. And that's going to be the solution if they don't move. But I think it would be better if they did it. It would certainly be more secure and easier to set up and use. I don't know, maybe Hector will do this with, uh, with the right tool or something. Uh, Hector Garcia. All right, David, that's all the time we have this week. Um, don't forget, listeners, you can get a free continuing professional education credit for listening to the accounting podcast and all and all of our past episodes on the earmark app. Just go to Earmark app in your web browser, or download the free mobile app from the App Store. Create your free account and earn your free CPE. It's free. One per week. And if you want to support us and unlock unlimited courses, it's just 130 $150 a year. That's it. It's the best deal on the market. We've got 2000 courses now on the app, and you can listen to a podcast episode while you're driving to work, while you are working out, while you're doing chores. And when you're done listening, you just take a quick quiz and you get your CP certificate right there and it gets emailed to you as well.

David Leary: [01:09:59] So I was looking like from a from a value perspective, I was looking at some continuing education apps that were in other industries. Nobody's everybody is like hundreds and hundreds of dollars a year. Like we're giving you unlimited for Limited for $150 a year. You can't beat it. And we have way more courses.

Blake Oliver: [01:10:13] That's right. Well, and the reason we keep it so cheap, we could raise the price, but the reason we keep it affordable is because we want to make good, interesting CPE available to everyone. It shouldn't just be like partners who can afford to spend 1000 or $2000 on CPE every year, and you know, everybody should get to do it. You shouldn't have to go sit on like, free webinars that have nothing to do with what you do just to get a CPE. Like listen to a show you like. And if you don't like our show, there's tons of other shows. Um, on the on on the app, including oh My fraud.

David Leary: [01:10:50] But if you're still listening to this, you must like our show or it's a hate listen for you and you just.

Blake Oliver: [01:10:53] Or you just dropped in randomly on LinkedIn or Twitter and you're like, oh, this is terrible. Well, I'm telling you, there's even better content on the earmark app. Oh, my fraud is very popular, as is federal tax updates. There's also my earmark podcast. We got so many more on the app. Go check it out. Earmarked app, free account. Sign up and we will see you all here next week. Um, actually I'm going. I don't know when we're going to record next, David, because I'm taking a little vacation, so I won't be here on Wednesday, but we'll figure it out.

David Leary: [01:11:27] Are you skiing or something?

Blake Oliver: [01:11:29] I'm going to be, uh. Yeah, I'm going to be skiing.

David Leary: [01:11:32] Uh, don't come back with another broken hand or thumb or something.

Blake Oliver: [01:11:35] I'm gonna wear my brace. Don't worry about it. I'll be good. All right. Um, I'll be safe. I'm only going to stick to, like, the greens and blues. No, no, no, no black diamonds for me this trip.

David Leary: [01:11:47] Okay, well, make sure we have to apple tag you or whatever.

Blake Oliver: [01:11:50] Or tag me.

David Leary: [01:11:52] To make sure you're only on the blues and green diamonds give you a special vest that says you're only allowed to do these.

Blake Oliver: [01:11:57] Safety vest.

David Leary: [01:11:58] Yeah. Safety vest.

Blake Oliver: [01:12:00] Awesome. Bye, everyone. See you around.

Speaker3: [01:12:02] Bye.